The ultimate goal of this research program is to identify brain biomarkers of Autism Spectrum Disorders (ASDs) and biologically-based ASD subtypes using multimodal Magnetic Resonance Imaging (MRI). Previous research into the neurobiology of ASD provides evidence for anatomical and functional disturbances across multiple neural systems that correlate with ASD symptom severity. Our research program builds on these findings by showing how disturbances in different neural systems in ASD correspond to unique profiles of behavioral and cognitive impairment. Thus far, 38 children and adults with ASD and 39 healthy controls, group-matched by age, sex, race, intelligence quotient (IQ), socio-economic status, and handedness, have been enrolled (targeted enrollment 100 ASD and 100 control subjects). Phenotypic profiling includes gold standard clinical instruments to assess the severity of impairment in the core behavioral domains of ASD along with measures of intelligence, procedural and declarative memory, visuospatial processing, and executive functioning. Participants undergo high-resolution anatomical MRI, functional MRI (fMRI), diffusion tensor imaging (DTI), and magnetic resonance spectroscopy (MRS) in a 3 Tesla scanner. Anatomical measures include fine-grained measures of local volumes of the cortical surface, the underlying white matter, and subcortical gray matter nuclei. DTI provides information about the orientation and integrity of white matter fiber tracts. Multiplanar Chemical Shift Imaging (MPCSI) provides extremely high quality measures of metabolites, including measures of neuronal density, in small contiguous voxels throughout the brain. Brain activation during a task of emotional face recognition is measured using fMRI in conjunction with eye-tracking. We analyze these diverse forms of data within a single imaging space using advanced image acquisition, image processing, and statistical modeling. These methods enable the precise detection of inter-individual and inter-group variability at each voxel, within and across modalities, permitting a deeper understanding of the relationships between local volumes, functional activity, metabolite concentrations, and connectivity. We hypothesize that neurobiological ASD subtypes will emerge based on differential patterns of disturbances in distinct neural systems that have been shown to be altered in ASD, such as amygdala-hippocampal, frontostriatal, and the mirror neuron systems, among others. We have previously defined neuroanatomical signatures for Tourette Syndrome1, 2, Attention-Deficit/Hyperactivity Disorder3, 4, familial depression5, adult schizophrenia6, and prematurely born children.7 We have also developed methods that enable the classification of neuropsychiatric disorders and their subtypes based on detailed measures of neuroanatomy. Using these techniques, our research program will define brain biomarkers for ASD, biologically-based ASD subtypes, and the neural bases for the phenotypic heterogeneity of these complex disorders. PUBLIC HEALTH RELEVANCE: Autism Spectrum Disorders (ASDs), affecting an estimated 1 in 150 individuals, are complex disorders of brain development that cause lifelong impairments in social ability, communication and behavioral flexibility with high rates of intellectual disability and medical comorbidities. The identification of brain-based biomarkers in individuals with ASD will illuminate the neural bases of this heterogeneous disorder and identify biological subtypes of ASD thereby yielding enormous benefits in the search for vulnerability genes for ASDs. Our proposed research plan will advance public health by enabling strategies for prevention, early detection, and personalized treatment of ASDs.